E S T I M A T I O N OF FECAL OUTPUT VICTOR A. LAZAR and WARREN G. MONSON
X-Ray Spectrographic Determination of Chromic Oxide in Steer Feces
Agricultural Research Service, U. S. Department of Agriculture, Ithaca, N. Y.
The x-ray spectrographic procedure is based upon the linear relationship of the counting rate at the clhromium K, line to the chromic oxide concentration. The working curve i s based on the chemically determined chromic oxide concentration in selected fecal samples. The method i s useful for the routine rapid determination of chromic oxide.
C
oxide has gained widespread use as an external fecal indicator since it was first suggested by Edin in 1918 (7). I t is one of several indicators which may be employed to estimate the fecal output of animals in feeding and grazing trials. Its use for this purpose has been adequately reviewed by Reid (70). Several procedures for the chemical determination of chromic oxide in feeds and feces are available. Colorimetric ( 7 . 3, 7 7 ) . volumetric (E), and atomic absorption (74) methods, among others, have been described. Some of these have been compared by Stevenson and De Langen (72). Methods currently available are timeconsuming and some of them present inherent difficulties in obtaining reproducible results (5. 72). This paper reports an x-ray spectrographic technique for the determinati'on of the chromic oxide concentration in feces of grazing steers. HROMIC
Equipment A Philips standard x-ray spectrograph \vith full \\-ave rectification and current stabilization \vas used. A tungsten-target F.4-60 x-ray tube was operated a t 19 kv. and 18 ma. A lithium fluoride crystal. 2d spacing of 4.0276 A , . \vas used in a n air path. The collimation system consisted of one 4- X '/8-inch and one 4- X 0.005-inch Soller slit assemblies. T h e scintillation detector. Ivith a thallium-activated sodium iodide crystal, W E I S operated at 1050 volts. The modified sample holder and the Mylar-backed aluminum alloy frames have been described (2).
fed 1 pound of pelleted corn meal containing 20 grams of chromic oxide every morning. The chromic oxide was added to the corn meal in the form of finely ground chromic oxide "bread" prepared in a manner similar to that described by Kane, Jacobson, and Moore (8). The \\-orking curve is based upon selected samples of feces which Ivere analyzed chemically for chromic oxide according to the colorimetric method of Brisson ( 3 ) . The curve is obtained by plotting the counting rate a t the chromium K , line (X = 2.291 A , 28 = 69.34') as a function of the chromic oxide concentration in the samples. The counting rate, based on the time required to record 25,600 events: was found to be directly proportional to the concentration of chromic oxide. The equation of the line for this relationship was calculated by the method of least squares ( 4 ) . The concentration of chromic oxide in the unknowns lvas then calculated from the above equation. (The concentration of chromic oxide may also be read directly from the working curve.)
Results and Discussion X-ray spectrographic data may be related to the concentration of a n element in a given sample by several means. Three of these ii-ere used to relate data obtained in this study to the concentration of chromic oxide in the fecal samples.
Sleasure F \vas found to be as good as either F - S or F,'S (Table I). Studies already reported sho\ved that the ratio of the counting rate a t the K , line to the counting rate at background \vas a good measure of the concentration of an element in the parts per million range (2. 9. 73J. The chromic oxide concentration in the samples being analVzed. hov.ever. was expected to be between 0.2 and 1.0Tc. \Vith this chromic oxide concentration, a relatively low po\ver settino; gave the desired counting rate a t the chromium K , line Ivhile maintainin? nearly constant the counting rate at the background. The feasibility of utilizing synthetic reference samples to develop a \\-orking curve \\-as investigated. Stock s p t h e t i c reference samples of feces containing 1.00 and 2.005 chromic oxide \\-ere prepared. and portions ivere added to feces free of chromic oxide to provide samplc.; \\ith a range of 0.00 to 1.207,. The cllr\rei based on this set of synthetic reference samples
Table 1. Comparison of Three Possible Measures of Chromic Oxide Concentration Based on Chemical Analysis of 30 Samples
Procedure The samples were oven-dried at 70' C., ground to pass a 20-mesh screen, and stored in capped bottles. T h e determinations were made on about 0.5-gram subsamples evenly spread in the frames described. As previously reported, it was not necessary to pack weighed quantities of the samples into the frames (2). The feces samples were collected tlvice daily according to the experimental design and with the cooperation of the Holstein-Friesian steers that were grazing on experimental alfalfa-grass pastures. The animals were
F. The counting rate a t the chromium I;, line. F - S. The counting rate at the chromium K , line less the counting rate a t background ( A = 2.164 A . 28 = 65.00O). F'S. The ratio of the counting rate at the chromium K , line to the counting rate a t the background.
Measure
Fc
Correlafion wifh Chemicol Deferminafiona
Constants o f W o r k i n g Curves
F = 2 4 . 6 f 227.5xd F - S = 4.1 226.8~F / S = 1 ,240 I O . 726x FIS Correlation coefficient r = Zxy .\/Fx2Z7 ( 6 ) .
F-
Se
+ +
M e a n and Standard Deviofion from Working Curve
0.997 0.997 0.995
0.019 i. 0.021 0.019 i 0.021 0.025k 0 022
Q
1xr
-)
Mean ( 2 = T:) with standard deviation (J = 1 -(6)of absolute differn-1 ences between concentrations determined chemically and calculated from equations of working curves. F.Counting rate. c.p.s., at chromium R, line, X = 2.291 A , 28 = 69.34'. d x . Chromic oxide concentration ranging from 0.01 to 1.19 yc. S. Counting rate. c.P.s., at background, X = 2.164 h.,28 = 65.00 '.
VOL.
1 3 , NO. 2 M A R . - A P R .
1965
161
Table II.
= +
Constants for Equations of lines ( y 6 mx) and Correlation Coefficients for Five Sets of Reference Samples F
Base of Reference Sample
Chemical determination Dilution of stock chromic oxide synthetic reference samples Addition of Chromic oxide powder Sodium dichromate solution Chromic oxide bread
b
m
24.6
227.5
0,997
b
4.1
Correlafion coeff.
m
b
Correlation coeff.
m
Chromic Oxide Concentrations,
72 2 2 6 . 8 0.997
Samples
1 , 2 4 0 10.726 0.995
Prepared by
24.6
279.9
0.999
3.4
278.5
0.999
1 . 2 3 3 12.492 0.997
2 9 . 2 267.2
0,998
7.3
268.7
27.3
283.8
0.999
6.2
283.3 0.999
1 . 4 4 4 12.911 0.998
27.5
165.1 0.999
6.5
165.7 0.999
1.271
had the same intercepts as the curves based upon the chemically analyzed samples. but their slopes \\.ere appreciably steeper (Table 11). T h e curves based upon synthetic reference samples prepared by adding increments of chromic oxide (finer than 80-mesh) to Tveighed portions of dried ground feces ivere similar to the curves based on the first set of synthetic reference samples. Furthermore. curves based upon synthetic reference samples prepared by adding equivalent amounts of chromium as an aqueous solution of sodium dichromate to dried ground feces previously moistened with acetone were similar to the above curves. A set of curves ivith markedly different slopes, however, was obtained when increments of the chromic oxide "bread" were added to portions of dried ground feces. The curves based on this set of synthetic reference samples had about the same intercepts for all three measures as those obtained on the other sets of reference samples, but they had slopes Ivhich Ivere considerably lo\ver than the slopes of the other curves. A study of these curves indicates that the addition of chromium as po\vdered chromic oxide, as an aqueous solution of sodium dichromate, or as chromic oxide bread results in a set of synthetic reference samples whose curves differ markedly from those based on the chemically determined chromic oxide concentration in selected samples (Table 11). The use of the curves based upon synthetic reference samples resulted in in the chromic differences of 20 to oxide concentrations. The magnitude of these differences led the authors to check chemically the chromic oxide concentrations in the synthetic reference samples. Chemical chromic oxide determinations (3) on selected synthetic reference samples showed that there \vas little possibility of major error in the adding and mixing
162
FIS
F - S Correlation coeff.
Table 111. Comparison of Calculated and Duplicate Chemically Determined Chromic Oxide Concentrations in Selected Synthetic Reference Sample
J. A G R . F O O D CHEM.
0 , 9 9 8 1.095
13.384 0.998
8 . 1 8 3 0.999
of chromic oxide or its equivalent as an aqueous solution of sodium dichromate with portions of the dried ground feces. The chemically determined chromic oxide concentrations were Lvithin Sq;i, of the calculated values (Table 111). A study of the reproducibility of the chromic oxide determinations indicated that the error due to subsampling was about the same as the error expected in counting 25,600 events ivhen measure F or F - S was used. The error due to subsampling was about three times the error expected in counting 25,600 events ivhen measure F/'S was used. The error in the determination of the chromic oxide concentration based upon single observations on each of ten subsamples was slightly higher than that based upon ten observations on a given subsample which had not been removed from the sample chamber benveen observations (Table IV). The discrepancies in the slopes of the jvorking curves based on the synthetic reference samples \\.hen compared to the slopes of the working curves based on the set of selected chemically analyzed samples are difficult to explain. From considerations of the path followed by the feed concentrates containing chro~
~
~
~
Dilution of chromic oxide stock Addition of Powdered chromic oxide Sodium dichromate solution Chromic oxide bread
Chemical Calcd.
I
II
0.50
0.47
0.46
0.50
0.47
0.48
0.50
0.48
0.48
0.80
0.75
0.80
mic oxide and the roughages obtained by grazing, it would be logical to assume that the digestive process of the steers is a major factor. The authors have been unable to duplicate the effects of this process in vitro. I t is obvious from the results tabulated in Table I1 that the addition of chromic oxide as a powder. as the chromic oxide bread, or as the equivalent amount of sodium dichromate in an aqueous solution to portions of dried ground feces \vi11 not duplicate the state of the chromic oxide in the samples collected from grazing steers. Furthermore. the counting rates obtained on a small number of chemically analyzed excreta of chicks fed a semipurified diet containing chromic oxide differed markedly from the counting rates obtained on chemically analyzed (3) fecal samples from grazing steers.
Conclusions T h r x-ray spectrographic technique described is rapid (as many as 10 determinations an hour), requires a minimum of sample pretreatment, and is sufficiently accurate for all but the most exacting kind of work. However, the x-ray spectrographic data obtained
~~~
Table IV. Means and Standard Deviations. of Chromic Oxide Concentration Based on Single Observations on Each of 10 Subsamples and 10 Observations on One Subsample No. Sam- of ple SubNo. sampler
1
10 1
2
10
No. o f Observafionr
1 10 1 10
1 a
Measure
F - S =k 0.007
F
0.476 0.468 0.748 0.743
- = -.zx
Mean: X
Standard deviation:
s =
0.477 0.471 0.750 0.745
i 0.007 f 0.005 i 0.010 f 0.006
d7 ~
n - 1
(6).
i 0.005 f 0.008 f 0.006
FIS
0.491 i 0.011 0.490 f 0.017 0.770 0.008 0.762 f 0.020
+
on samples being analyzed must be compared to x-ray spectrographic data obtained on similar refee.rence samples analyzed chemically for chromic oxide. Furthermore. the results of a limited study showed that x-ray :spectrographic data obtained on reference samples of feces from grazing steers differed markedly from those obtained on samples of excreta from chicks fed .I semipurified diet containing chromic oxide.
Literature Cited (1) Bolin, D. LY., King, R. P. Klosterman. E. W.. Science 116,634 (1952). (2) Brandt, C. S., Lazar, \‘. .4., J. A G R . FOOD CHEM. 6,306 (1958). 13) Brisson. G . J. Can. J . Aer. - Sci. 36,210 (1956). (4) Crumpler, T. B., Yoe, J. H., “Chemical ComDutations and Errors,” p. 220, M’iley Sew York, 1940. (5) Day, K . M., Science 120,717 (1954). ( 6 ) Dixon. W. J.. Massey, F. J., Jr., “Introduction to Statistical Analysis,” p. 20) p. 163, McGraw-Hill, New York, 1951. (7) Edin, H.. .Medd. Centralanst. Forsoksaasendet Jordbruks. 105,l (1918). ( 8 ) Kane. E. -4.: Jacobson, \V. O., hfoore, L. A . . J . .\‘zit. 41, 583 (1950). (9) Lazar, V. A , , Beeson, K . C., J . j
Acknowledgment Samples of chick excreta and their respective chromic oxide concentrations \\ere supplied by Jimmy D . Garlich, Department of Poultry Husbandry. Corne11 Universit);, Ithaca. 5‘.Y.
,
Assoc. O@c. Agr. Chemists 41, 416 11958). (IO) Re’id, J. T.: “Pasture and Range Research Techniques,” pp. 51-6, Comstock Publishing Associates, Ithaca, S.Y., 1962. (11) Schurch, A. F., Lloyd, L. E., Crampton, E. W. J . .l’utr. 41, 629 ( 1950) . (12) Stevenson, ,4.E.: De Langen: H., .Yere Zealand F. ‘4gr. Res. 3, 314 (1960). (13) Whittig: L. D.? Buchanan. J. R., Brown. .4. L., J. AGR. FOODCHEM. 8,419 (1960). (14) \Yilliams. C. H.. David. D. J., Iismaa, 0.: J . ilgr. Sci. 59, 381 (196 2). Received f o r reciew June 8, 75’64. Acctpted October 8. 7g64.
P L A N T U P T A K E OF M I N E R A L S NELSON 0 . PRICE and W. W. MOSCHLER Virginia Polytechnic Institute, Blacksburg, Va.
Effect of Residual Lime in Soil on Minor Elements in Plants
Residual lime in the soil resulted in significant changes of the plant composition of peanut foliage, soybean foliage, and orchard grass 7 and 9 years after lime application. Copper, cobalt, manganese, iron, and zinc were significantly lowered in each plant. Molybdenum was significantly increased.
L
has long been used as a fertilizer and for correcting a.nd controlling soil acidity. The effect of soil acid conditions can be observed by examining plants gro\vn on soils \Lhich have incremental dressings of lirne. One such study. \\hich illlxstrates the depressing effects of lime on the levels of cobalt. nickel, and manganese i n pasture herbage and its enhancing effect upon molybdenum. was conducted b\- Mitchell (9). H e used the equivalent of 6 and 12 tons of C a C 0 3 per acre wiih the p H of the soils at 5.4 for control and 6.1 and 6.4. respectively. for the levels of CaC03. He found marked depression in cobalt, nickel. and manganese content Lvith decided increases in molybdenum. Taylor (75) and coworkers found that lime significantly decreased the manganese content of plants, significantly increased the molybdenum content, but normally has little influence on the level of copper. Fiskell (7) showed that liming reduced intake of manganese content of blue lupines after 4 years’ application and that symptoms of a copper deficiency in ladino clover were intensified by a heavy liming program. Robinson’s (7.4) \vork under greenhouse studies shoived that the molybdenum content of alfalfa, crimson clover, and .Austrian wini;er peas groivn on acid soils high in molybdenum may be increased by heavy liming to a point IME
\\here these crops are toxic IO cattle. Other workers ( 7 , 2, 4. 8:7 7 . 73, 76, 77) confirm the effects of soil liming on the composition of plants and the uptake of minerals. The capacity of pastures to supply the mineral needs of grazing stock in adequate amounts: in proper proportions, and without harmful excess depends upon many interrelated factors. Controlled experiments (70) in 1953 and 1954 over wide areas of the state provided opportunity to study residual lime effects for periods u p to 7 or 9 years after application. T h e literature does not report the effects of soil lime treatments on the minor element uptake of plants after years of application.
Procedures Samples Ivere collected from lime plots in three areas of the state: peanut tissues (leaves) from Norfolk loam fine sandy, deep phase, near the Holland Experiment Station; soybean tissues (leaves) from Sassafras sandy loam, near the LVarsaw Experiment Station; and orchard grass from Lodi loam, near Abingdon, Va. The peanut and soybean tissues \\ere collected 7 years and the orchard grass 9 years after lime treatment. The experimental sites \\ere located on private farms. After the lime was applied, the oivners continued to crop the areas with normal VOL.
practices, except that no additional lime \vas applied. Fertilization varied, but in most cases the rates used were less than those recommended; and it is believed that neither crops nor fertilization materially influenced p H or base relationships in the soil. No crop yield response was noted except in the case of soybeans, where approximately a 25y0 increase \vas obtained from the use of lime compared to no lime. KO grazing was done on the peanut or soybean soils, but the orchard grass was in continuous pasture for dairy cattle. hlineralogically. the three soils were composed primarily of vermiculite 10 75yc) and contained smaller quantities of illite and kalonite (10 to 25oc,). ‘Total exchangeable cations present in the unlimed soils were 5.6. 5.9, and 9.6 meq. per 100 grams of soil for the peanut. boybean, and orchard grass soils. re>pectively. Soil p H \vas determined on a Model G: Beckman pH meter. using a 1 to 1 soil-Ivater ratio and a half-hour equilibration period. Other soil properties are discussed fully b>- hioschler /r 70). Cobalt. copper, iron. manganese. and zinc \\ere determined according to official merhods of the .40.4C ( 3 ) . Molybdenum was determined by the thiocyanate stannous chloride method of Evans (6). The analyses are reported on a dry \\eight basis, and wet digestion (nitric and perchloric acids) was used instead of dry ashing.
1 3 , NO. 2 M A R . - A P R .
1965
163
